Scholarlink

<p>To reduce the economic loss caused by data damage and loss of power system failure, this paper, it’s on account of the non-single objective granule majorization arithmetic, which includes the disaster recovery service layer, virtualized resource management layer, resource layer, distributed storage layer, and storage node layer. Based on the multi-objective particle swarm optimization algorithm, this architecture follows the principles of minimizing data storage costs, minimizing data recovery time, and maximizing system availability, to achieve distributed storage and rapid recovery of power dispatching data. To enhance the algorithm's performance, disaster recovery effectiveness, and backup capability for power dispatch data, several key mechanisms were incorporated into the multi-objective optimization algorithm. These include the particle position update and speed adjustment strategy, non-dominated sorting with Pareto frontier solution selection, a dynamic weight adjustment mechanism, and a balance between global and local search capabilities. The experimental results demonstrate the effectiveness of the proposed strategy. It achieves a node distribution rate of up to 90%. The data recovery times are 9&#xa0;s for the central node, and 1780s, 1790s, and 1780s for storage nodes 1, 2, and 3, respectively. Furthermore, the maximum data backup time across all backup sources is 9.32&#xa0;s, with a peak backup space occupancy rate of only 0.0007%. Throughout the tests, the power dispatch performance was consistently maintained above 98%. These results indicate that the proposed multi-objective disaster recovery strategy for power dispatch data possesses significant practical application value.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Disaster recovery strategy of power dispatching data based on multi-target particle swarm optimization algorithm

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年4月17日

查看全文
  • QianQian Xu,
  • Jing Gao,
  • Jian Li

摘要

To reduce the economic loss caused by data damage and loss of power system failure, this paper, it’s on account of the non-single objective granule majorization arithmetic, which includes the disaster recovery service layer, virtualized resource management layer, resource layer, distributed storage layer, and storage node layer. Based on the multi-objective particle swarm optimization algorithm, this architecture follows the principles of minimizing data storage costs, minimizing data recovery time, and maximizing system availability, to achieve distributed storage and rapid recovery of power dispatching data. To enhance the algorithm's performance, disaster recovery effectiveness, and backup capability for power dispatch data, several key mechanisms were incorporated into the multi-objective optimization algorithm. These include the particle position update and speed adjustment strategy, non-dominated sorting with Pareto frontier solution selection, a dynamic weight adjustment mechanism, and a balance between global and local search capabilities. The experimental results demonstrate the effectiveness of the proposed strategy. It achieves a node distribution rate of up to 90%. The data recovery times are 9 s for the central node, and 1780s, 1790s, and 1780s for storage nodes 1, 2, and 3, respectively. Furthermore, the maximum data backup time across all backup sources is 9.32 s, with a peak backup space occupancy rate of only 0.0007%. Throughout the tests, the power dispatch performance was consistently maintained above 98%. These results indicate that the proposed multi-objective disaster recovery strategy for power dispatch data possesses significant practical application value.